🔒 pin tests.yml actions to commit SHAs (#721)

Update README.md

Update citation (#688)

Bump vLLM and TRL (#665)

Fix Weka refresh (#666)

Set DP=2 for smol model evals (#664)

Refresh Weka on Slurm (#662)

Align EOS token ID between tokenizer and generation config (#663)

Bump deps (#656)

Add OpenR1-Distill recipe (#661)

Add better logging defaults for GRPO (#657)

GRPO with codeforces problems (#627)

Bump lower bound on liger-kernel (#654)

Bump DeepSpeed to 0.16.8 to fix OOM on Qwen3 (#653)

Add dataset mixer (#647)